lvs/tun模式下一台realserver的tunnel down后引发的问题请教

lvs/tun模式，HA+ldirectord，一台director带四台realserver,运行squid服务，发现realserver的tunnel down以后，hash表中该realserver的权值没变为0，(如果是realserver宕机或运行的服务down了，权值会变为0的），ldirectord.cf配置如下
# Global Directives
checktimeout=3
checkinterval=1
#fallback=127.0.0.1:80
autoreload=yes
logfile="/var/log/ldirectord.log"
#logfile="local0"
quiescent=yes

virtual=192.168.100.203:80
real=192.168.100.125:80 ipip
real=192.168.100.126:80 ipip
real=192.168.100.124:80 ipip
real=192.168.100.128:80 ipip
service=http
checkport=80
checktype=connect
scheduler=wlc
protocol=tcp
在客户端浏览器中输入192.168.100.203，就会显示找不到服务器，这里只是一台realserver的tunnel down了而已，其他的realserver都是正常的，为什么会是这样的结果啊？假如在tunnel down掉的那台realserver上运行telnet 192.168.100.203 80,就会出现如下结果
[root@CVS init.d]# telnet 192.168.100.203 80
Trying 192.168.100.203...
Connected to 192.168.100.203 (192.168.100.203).
Escape character is '^]'.
说明是连接上了，在此同时用netstat -na|grep 192.168.100.203时却是这样的
[root@CVS ~]# netstat -an|grep 192.168.100.203
tcp 0 0 192.168.100.203:32780 192.168.100.203:80 ESTABLISHED
tcp 0 0 ::ffff:192.168.100.203:80 ::ffff:192.168.100.20:32780 ESTABLISHED
192.168.100.20这个ip是ping不通的，不知道是哪里出了问题，谢谢！